Research On The Regulation Of "Rat Trading" Behavior Based On Game Theory

This paper summarizes the underlying securities regulatory theory, game theory and other related literature. By establishing two game models between regulatory authority and fund managers, we study the optimal game strategy of both sides and analyze the basic reasons of the failures to prohibit ’rat trading’. Based on the result we give some policy suggestions. Besides, we make a quantitative research of some aspects, such as price, excess returns and abnormal volatility, of the stocks involved in’rat trading’according to disclosed details.The main contents are as follows:First, we describe the development history of China’s stock market and fund industry, and summarize the basic features of "rat trading" and some related research in this field, including securities regulatory theory, game theory, and some classic literature of’rat trading’behavior.Second, we construct a static game model of complete information with relatively relaxed assumptions. In this model, there are two kinds of regulations, active regulation with high-cost and passive regulation with low-cost. The former can hunt down the ’rat trading’ but the later cannot. The equilibrium is studied with evolutionary game theoretical models. Then, we build a much more complex model by assuming the probabilities of hunting down the ’rat trading’for active regulation is a, while the passive regulation is β. We use the same method to depict the entire game process.Third,we amend the assumptions to construct a dynamic game model with incomplete information. In the real word,regulatory decisions on both sides are generally not occur simultaneously.We believe that the different regulatory authorities will have a large gap in the quality of human resourca and technical levels, so we distinguish for the high-level and low-level of regulatory authorities, and accounting for respectively p and 1-p. We assume that the high-level regulatory authoritiy can find out the "rat trading" using an active monitoring strategy; the high-level regulatory authoritiy can find out the "rat trading" with a probability of a with an negative monitoring strategy; the low-level regulatory authoritiy can find out the "rat trading" with a probability of β with an active monitoring strategy;and the low-level regulatory authoritiy can not find out the "rat trading" using an negative monitoring strategy. Under the above assumptions, we discuss eight different game results depending on the different parameters.At last, we make several empirical tests of ’rat trading’in China’s stock market.24 samples that are already disclosed are chosen. First, we study the price features of these stocks by comparing the mean value, variance and normality assumption of prices before, in and after the ’rat trading’. Second, we use event research method to test the excess return of these stocks by comparing the AR, AAR and CAR value. Third, we use GARCH model to test the abnormal volatility.